1. Yurii Levashov Undula t or fiducialization test resultsylevash@slac.stanford.edu Oct. 14, 2004 Undulator Fiducialization Test Results Fiducialization Tolerances and Procedure Pointed Magnet Calibration Two - Point Measurement Algorithm Hall Probe Center Calibration Results Summary

2. Yurii Levashov Undula t or fiducialization test resultsylevash@slac.stanford.edu Oct. 14, 2004 2 Fiducialization Tolerances Fiducialization errors are small parts of alignment ones. Magnetic measurement bench will be used for LCLS undulators fiducialization. ParameterAlignment ValueFiducialization ValueUnits Horizontal location tolerance25030 µm Vertical location tolerance5010µm Longitudinal location tolerance50050µm * - “Magnetic Measurements and Alignment at SLAC” by R. Ruland and Z.Wolf. DOE Review, Aug. 10-12, 2004. Table 1 *

3. Yurii Levashov Undula t or fiducialization test resultsylevash@slac.stanford.edu Oct. 14, 2004 3 Undulator with Pointed Magnets Extra block of 25 cm in length is between the pointed magnet fixture and the undulator. The block attached flat to the undulator end plate by screws. A “tooling ball-bushing” pair is used to locate the magnets with 200µm accuracy. Position of beam axis with respect to the fiducials    p.magnet     

4. Yurii Levashov Undula t or fiducialization test resultsylevash@slac.stanford.edu Oct. 14, 2004 4 Fiducialization Procedure Calibrate the pointed magnets on a special stand. Measure magnetic center to tooling ball offset (    Measure and tune the undulator. Move the Hall probe into a position which gives a desirable K value. Move the probe along the ideal beam axis to the center of the magnets attached to the undulator. Find the magnetic center of the pointed magnets (zero field point). Measure offset between the undulator axis and the magnetic center (  p.magnet ). Measure the position of the pointed magnet tooling balls with respect to the undulator ones on CMM (    Apply the offsets to the ideal beam axis.

5. Yurii Levashov Undula t or fiducialization test resultsylevash@slac.stanford.edu Oct. 14, 2004 5 Pointed Magnet Calibration Magnets are placed on a calibration fixture. Hall probe is placed at the magnetic center Y 1. Magnets are rotated 180˚. Hall probe is placed at magnetic center Y 2.  is a difference in the Hall probe positions. D is measured on CMM. Assumption: Central point will not change its location. Question: How accurate could the magnetic center be found?

6. Yurii Levashov Undula t or fiducialization test resultsylevash@slac.stanford.edu Oct. 14, 2004 6 Pointed Magnet Calibration Fixture Magnets are at 180˚ Kinematics’ mount: 3 tooling balls instead of a cap. Two dowel pins instead of V – block 2 flat resting pads. Advantages: Very high surface quality On-shelf parts Hall probe fixed on a support table. Magnets are moved by X,Y,Z stages. Reliable measurement algorithm has been implemented

7. Yurii Levashov Undula t or fiducialization test resultsylevash@slac.stanford.edu Oct. 14, 2004 7 Why Magnets are not at 90˚? Magnetic field from one pointed magnet extends far enough to interfere with one from the second. There is no zero field point. Magnets should have smaller sizes and be set apart at a bigger distance. It involves a different design. Don’t have a reliable algorithm to implement the 90˚ scheme. Will continue work on this issue.

8. Yurii Levashov Undula t or fiducialization test resultsylevash@slac.stanford.edu Oct. 14, 2004 8 Test Measurement Set-Up Difference in thermal coefficients causes in Y positioning errors: ~ 5µm / ˚C. ( Will be not a case in new Lab.) Magnets misalignment: ~ 300µm in X and ~ 150µm in Z.

9. Yurii Levashov Undula t or fiducialization test resultsylevash@slac.stanford.edu Oct. 14, 2004 9 Two - Point Measurement Algorithm 1.Start position is unknown. It is randomly chosen with ± 200 µm (cube). 2.Move in Y by 1.8 mm toward the first magnet. 3.Scan in Z, then in X. 4.Find maximums by polynomial interpolation. 5.Move back in Y by 1.8mm. 6.Scan in Y. 7.Find a minimum by linear fit. 8.Repeat steps 2 – 7 for the second magnet. 9.Calculate magnetic center position. Advantage: Information about misalignment of the magnets.

10. Yurii Levashov Undula t or fiducialization test resultsylevash@slac.stanford.edu Oct. 14, 2004 10 Hall Probe Center Calibration Results Group 3 Hall Probe # 01231177 Std. X = 5µm Std. Y < 1µm Zeroes in Z (Stage resolution)  T ≈ 0.1 ˚C ≈ 4.5 minutes/run

11. Yurii Levashov Undula t or fiducialization test resultsylevash@slac.stanford.edu Oct. 14, 2004 11 Temperature Effect on Y Measurements Group 3 Hall Probe # 01231177 Std. X = 3.4 µm Std. Y < 1 µm (Trend)  T = 1.2 ˚ C Slope of 6 µm / ˚C is in a good agreement with our estimation. No temperature effect on X positioning

12. Yurii Levashov Undula t or fiducialization test resultsylevash@slac.stanford.edu Oct. 14, 2004 12 Effect of Steel Plate on Magnetic Center Location 18cm x 18cm steel plate Group 3 Hall Probe # 01231177 Magnets at 180˚ Minimum distance - 25cm

13. Yurii Levashov Undula t or fiducialization test resultsylevash@slac.stanford.edu Oct. 14, 2004 13 Summary Fiducialization procedure is close to be finalized. Calibration fixture based on pointed magnets technique has been developed. Work on the calibration fixture will be done soon. Two-point calibration algorithm has been proposed and tested to be adequate. Magnet misalignment of ± 200 µm has not affect on results of the calibration. Accuracy of locating the pointed magnet center is: 1 µm in Y; 5 µm in X; 20 µm in Z (stage resolution).